A user who clicks a button and sees a spinner for forty seconds has three thoughts in order: is it working, is it stuck, and should I try again. By the third thought they've either refreshed, losing the work, or decided the feature is unreliable.

Makes the Wait Worse	Makes the Wait Tolerable
A spinner with no information	A visible record of what it's doing, as it does it
A fake progress bar	Honest step counts: 'checking customs records (3 of ~6)'
No way to stop it	A stop button that actually works
Everything appears at the end	Findings appear as they're established
No sense of how long it takes	'This usually takes about 30 seconds'

Show the Work as It Happens

Streaming a running account of what the agent is doing converts dead waiting into something closer to watching a colleague work. It also does something more valuable: it lets the user notice early that the agent has misunderstood, and stop it, which saves the remaining forty seconds and, more importantly, preserves their trust.

Agentic AI and MCP for Product Managers

Module 1What 'Agentic' Actually Means

Module 2How an Agent Actually Works

Module 3MCP: What It Is and Why You Should Care

Module 4Designing the Tool Layer

Module 5Permissions and the Human in the Loop

Module 6Evaluating Agents

Module 7Cost, Latency and Unpredictability

Module 8Designing What the User Sees

Module 9Risk and Blast Radius

Module 10When an Agent Is the Wrong Answer

Module 11Final Project: Design One Real Agent End to End

Module 12Appendix: The Prompt Library

The Waiting Problem